939

I have a PHP script that can encode a PNG image to a Base64 string.

I'd like to do the same thing using JavaScript. I know how to open files, but I'm not sure how to do the encoding. I'm not used to working with binary data.

Kamil Kiełczewski
  • 53,729
  • 20
  • 259
  • 241
username
  • 15,721
  • 11
  • 37
  • 44
  • here is [another jquery plugin](http://archive.plugins.jquery.com/project/base64) fo base64 encode/decode – zahid9i Mar 14 '12 at 18:04
  • 2
    Here is best way to base64_encode and base64_decode using javascript. See below links. http://phpjs.org/functions/base64_encode:358 http://phpjs.org/functions/base64_decode:357 – gautamlakum Mar 28 '11 at 13:39
  • Check microjs: http://microjs.com/#base64 – Vinod Srivastav Mar 01 '16 at 14:16
  • Referenced in meta question *[Basically identical answers - Only difference: correction of errors](https://meta.stackoverflow.com/questions/371085)*. – Peter Mortensen Jul 16 '18 at 06:09

28 Answers28

1037

You can use btoa() and atob() to convert to and from base64 encoding.

There appears to be some confusion in the comments regarding what these functions accept/return, so…

  • btoa() accepts a “string” where each character represents an 8-bit byte – if you pass a string containing characters that can’t be represented in 8 bits, it will probably break. This isn’t a problem if you’re actually treating the string as a byte array, but if you’re trying to do something else then you’ll have to encode it first.

  • atob() returns a “string” where each character represents an 8-bit byte – that is, its value will be between 0 and 0xff. This does not mean it’s ASCII – presumably if you’re using this function at all, you expect to be working with binary data and not text.

See also:


Most comments here are outdated. You can probably use both btoa() and atob(), unless you support really outdated browsers.

Check here:

Lukas Liesis
  • 17,690
  • 7
  • 93
  • 89
Shog9
  • 146,212
  • 34
  • 221
  • 231
  • 49
    Note that this also works for webkit browsers, such as Safari. – Daniel Von Fange Sep 02 '10 at 09:59
  • 34
    Please note the special consideration for Unicode strings: https://developer.mozilla.org/En/DOM/Window.btoa#Unicode_Strings btoa and atob only work properly for ASCII based strings. As an American, you probably won't notice a difference ... but the first time you use an accented character, your code will break. – Dan Esparza Nov 08 '11 at 16:23
  • 88
    you should use `btoa(unescape(encodeURIComponent(str))))` if str is UFT8 – SET May 15 '13 at 07:49
  • 1
    If these functions support only ascii strings, then they should not be used at all. Period. You base 64 encode byte arrays, not character arrays (i.e. strings). Strings are logical character arrays, not byte arrays. Character arrays need to first be transformed into byte arrays by choosing a character encoding (UTF8,UTF16, or UTF32) to transform each character into a sequences of one or more bytes (for variable-length encoding schemes such as UTF8 or UFT16) or sequences of a fixed length (in constant-length encoding schemes such as UTF32). Once you've got a byte array, then base64 encode it. – Triynko Aug 06 '13 at 20:55
  • 5
    See my edit, @Triynko. These are not intended to be used to process *text*, period. – Shog9 Aug 06 '13 at 21:19
  • 1
    Ok, I was thrown off by Dan's comment about btoa and atob only working properly for ASCII strings. These function are, in fact, not meant to work with actual character strings at all, but rather with byte arrays that happen to be stored in strings, such that for each character in the string, calling charCodeAt(i) will always return a value between 0 and 255. I've seen this before in a Base64 JavaScript class that would first utf8 encode the string into another string (exactly as described where each "character" actually represents a byte value) and then base 64 encodes that. – Triynko Aug 06 '13 at 22:32
  • 1
    So an even more efficient method would be to use something like this (http://www.webtoolkit.info/javascript-base64.html) and just encode via: `var encoded = btoa(Base64._utf8_encode(input))` and decode using `var decoded = Base64._utf8_decode(atob(encoded))`. It seems to me that performing a utf8 encoding is much less error prone than passing the input through `unescape(encodeURIComponent(str))`. Also, the `string = string.replace(/\r\n/g,"\n");` line should be removed from the _utf8_encode function, since such a line break normalization has no business being in a basic utf8 encoding function. – Triynko Aug 06 '13 at 22:42
  • 4
    Here's a polyfill https://github.com/davidchambers/Base64.js , it's best to use the native functions and polyfill rather than include a library that introduces a new API. – Benjamin Gruenbaum Mar 18 '14 at 01:58
  • 2
    And the reverse to base64-decode a UTF8-string as encoded by the method of SET: decodeURIComponent(escape(window.atob(b64))); – Stefan Steiger Oct 28 '14 at 08:49
  • 2
    To whoever suggested [this edit](http://stackoverflow.com/review/suggested-edits/10172482): really cannot emphasize enough that you should not be passing arbitrary strings to `btoa()`. Yes, that example would work... But it's misleading; wide characters (legal in a string literal) would cause problems. – Shog9 Nov 11 '15 at 23:41
  • 4
    To better remember the names of the functions: `btoa()` is **b**efore **to** **a**fter, `atob()` is **a**fter **to** **b**efore – Dmitry Davydov Apr 10 '16 at 04:19
  • 4
    Javascript is such a drunk language. `atob` appears to stand for "alphanumeric to base64" so *of course* it actually does the opposite. –  Jul 14 '16 at 19:41
  • 4
    binary, @Will. The b stands for "binary". Like those sketchy newsgroups. – Shog9 Jul 14 '16 at 20:19
  • 4
    yeahbutstill... It's used to convert strings to base64... any non-drunk coder would have called it `toBase64`, would have supported unicode, and would have *then* gone out to drink. –  Jul 14 '16 at 20:31
  • 1
    So, that's where folks keep getting confused here, @Will - it absolutely should *not* be used to convert strings to or from anything, unless they're strings of Base64 (a.k.a, alt.binaries -> "b"). As you can see, the coders weren't drunk, just pervy. – Shog9 Jul 14 '16 at 20:40
  • 9
    It's pronounced `b to a` and `a to b`, b standing for binary and a standing for ASCII – Captain Hypertext Feb 03 '17 at 19:03
  • 1
    The "a" input/output should be valid ASCII, as that is the usual intent of Base64 @Captain. Do not under any circumstances assume that the "b" output is valid in any text encoding, nor send any text that doesn't fit into 8-bits into a function expecting "b". – Shog9 Feb 03 '17 at 19:08
  • 1
    @Shog9 As jpp [suggested](https://meta.stackoverflow.com/questions/371085/basically-identical-answers-only-difference-correction-of-errors/371114?noredirect=1#comment610921_371114), I'd like to bring my [meta question](https://meta.stackoverflow.com/questions/371085/basically-identical-answers-only-difference-correction-of-errors/) and corresponding [edit](https://stackoverflow.com/review/suggested-edits/20324632) of Sunny Milenov's answer to your attention. I hope that contacting you this way is acceptable. – T S Jul 17 '18 at 16:46
  • I'll try to find some time to look at that, @TS - but keep in mind, a lot of the answers here are doing extra stuff because, unlike the original asker, they're trying to pass Unicode *strings* through systems that don't support Unicode... And using a Base64 encoding as part of that process. While this is certainly understandable, it does make for some conflicting (and... ill-defined) requirements at times. – Shog9 Jul 17 '18 at 23:11
  • Not sure why this is an accepted or popular answer - it does nothing to address the Base64 specification by the OP. Instead, you're telling the OP to not use Base64 at all. There are other better answers which directly address the Base64 stipulation of the OP's question. – tpartee May 14 '21 at 17:20
  • Sure it does, @tpartee: the input/output of atob/btoa is base64. That's what the 'a' is in these abbreviations. That bit's easy enough, and I don't think too many people get tripped up anymore - what seems to throw them is the other side: the 'b'. Since originally JavaScript didn't have byte arrays, these functions are defined such that binary input must be contained in strings - but, and I keep having to emphasize this - they should NOT be passed arbitrary strings. That's not what they're for, or what they expect. – Shog9 May 14 '21 at 18:42
303

From here:

/**
*
*  Base64 encode / decode
*  http://www.webtoolkit.info/
*
**/
var Base64 = {

// private property
_keyStr : "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=",

// public method for encoding
encode : function (input) {
    var output = "";
    var chr1, chr2, chr3, enc1, enc2, enc3, enc4;
    var i = 0;

    input = Base64._utf8_encode(input);

    while (i < input.length) {

        chr1 = input.charCodeAt(i++);
        chr2 = input.charCodeAt(i++);
        chr3 = input.charCodeAt(i++);

        enc1 = chr1 >> 2;
        enc2 = ((chr1 & 3) << 4) | (chr2 >> 4);
        enc3 = ((chr2 & 15) << 2) | (chr3 >> 6);
        enc4 = chr3 & 63;

        if (isNaN(chr2)) {
            enc3 = enc4 = 64;
        } else if (isNaN(chr3)) {
            enc4 = 64;
        }

        output = output +
        this._keyStr.charAt(enc1) + this._keyStr.charAt(enc2) +
        this._keyStr.charAt(enc3) + this._keyStr.charAt(enc4);

    }

    return output;
},

// public method for decoding
decode : function (input) {
    var output = "";
    var chr1, chr2, chr3;
    var enc1, enc2, enc3, enc4;
    var i = 0;

    input = input.replace(/[^A-Za-z0-9\+\/\=]/g, "");

    while (i < input.length) {

        enc1 = this._keyStr.indexOf(input.charAt(i++));
        enc2 = this._keyStr.indexOf(input.charAt(i++));
        enc3 = this._keyStr.indexOf(input.charAt(i++));
        enc4 = this._keyStr.indexOf(input.charAt(i++));

        chr1 = (enc1 << 2) | (enc2 >> 4);
        chr2 = ((enc2 & 15) << 4) | (enc3 >> 2);
        chr3 = ((enc3 & 3) << 6) | enc4;

        output = output + String.fromCharCode(chr1);

        if (enc3 != 64) {
            output = output + String.fromCharCode(chr2);
        }
        if (enc4 != 64) {
            output = output + String.fromCharCode(chr3);
        }

    }

    output = Base64._utf8_decode(output);

    return output;

},

// private method for UTF-8 encoding
_utf8_encode : function (string) {
    string = string.replace(/\r\n/g,"\n");
    var utftext = "";

    for (var n = 0; n < string.length; n++) {

        var c = string.charCodeAt(n);

        if (c < 128) {
            utftext += String.fromCharCode(c);
        }
        else if((c > 127) && (c < 2048)) {
            utftext += String.fromCharCode((c >> 6) | 192);
            utftext += String.fromCharCode((c & 63) | 128);
        }
        else {
            utftext += String.fromCharCode((c >> 12) | 224);
            utftext += String.fromCharCode(((c >> 6) & 63) | 128);
            utftext += String.fromCharCode((c & 63) | 128);
        }

    }

    return utftext;
},

// private method for UTF-8 decoding
_utf8_decode : function (utftext) {
    var string = "";
    var i = 0;
    var c = c1 = c2 = 0;

    while ( i < utftext.length ) {

        c = utftext.charCodeAt(i);

        if (c < 128) {
            string += String.fromCharCode(c);
            i++;
        }
        else if((c > 191) && (c < 224)) {
            c2 = utftext.charCodeAt(i+1);
            string += String.fromCharCode(((c & 31) << 6) | (c2 & 63));
            i += 2;
        }
        else {
            c2 = utftext.charCodeAt(i+1);
            c3 = utftext.charCodeAt(i+2);
            string += String.fromCharCode(((c & 15) << 12) | ((c2 & 63) << 6) | (c3 & 63));
            i += 3;
        }

    }

    return string;
}

}

Also, search on "javascript base64 encoding" turns a lot of other options, the above was the first one.

Sunny Milenov
  • 20,782
  • 5
  • 75
  • 102
  • 3
    This is also useful when the base64 encoding is non-standard; in my case the "/" character wasn't used, and the "?" character was used instead, meaning even in Chrome atob() wasn't going to decode the base64 strings that were incoming. – Chris Moschini Dec 17 '11 at 12:43
  • 26
    Be careful with this code - it attempts to interpret your string as a UTF-8 encoded string. We had a case where we had a binary string (i.e. each character in the string should be interpreted as a byte), and this code did corrupt the data. Read the source, Luke. – Daniel Yankowsky Nov 12 '12 at 18:27
  • 2
    If you use code from webtoolkito info do not forget about copyright: /** * * Base64 encode / decode * http://www.webtoolkit.info/ * **/ – Maciej Łopaciński Dec 13 '12 at 11:36
  • 11
    All that is needed to make it safe for most binary encoding/decodings it to remove the questionable `string = string.replace(/\r\n/g,"\n");` statement in the utf8 encoding method. – Marius Jan 04 '13 at 17:59
  • I would use the above method with small amount of data. For bigger files is better to consider multi-part POST, where the file is transferred as binary. – Sunny Milenov May 06 '13 at 15:45
  • 7
    @Marius: I'm wondering why they would they even include `string = string.replace(/\r\n/g,"\n");` in the first place, lol. It's like "oh, lets encode this string, but first, why don't we just randomly normalize all the line breaks for no good reason at all". That should absolutely be removed from the class under all circumstances. – Triynko Aug 06 '13 at 22:40
  • I know Base64 was intended for other uses, and has a 33% overhead. Can this be optimized for web? So when you have an image with 100 Bytes, it gets saved in string form as 100 bytes instead of about 133 bytes? – employee-0 Aug 17 '13 at 22:04
  • 1
    @Triynko when people blindly implement solutions from here without checking it a little bit, then they probably get what they deserve. (not really, I don't mind helping people, but I can only hope they at least read these comments!) – Marius Sep 09 '13 at 12:55
  • 2
    I'm not a javascript guru, but this code seems to contain a bug: if chr2 is NaN, its value is still used in the statement `enc2 = ((chr1 & 3) << 4) | (chr2 >> 4);`. In my browser this works out ok, `NaN>>4` equals 0, but I don't know if all browsers do this (also, `NaN/16` equals NaN). – Jan Dec 16 '13 at 15:41
  • @Triynko: Today I discovered that I'd been affected by that line that clobbered "\r\n" into "\n". It certainly took me a while to suspect the base 64 encoding function of mangling my strings! – Johnny Kauffman Mar 18 '15 at 14:49
  • `input.charCodeAt(i++)` always returns 65535 for characters from 128(10000000)-255(11111111) in IE9! – Sen Jacob Jun 18 '15 at 03:00
  • Forcing an entire community to use linear complexity instead of constant (in decode) is a pretty unorthodox thing ;) – Géza Török Sep 12 '15 at 09:58
  • add .toString() to replace in _utf8_encode function because if json string is passed it will trigger exception – user889030 Mar 26 '20 at 07:18
  • Doesn't work on unicode. Do: Base64.encode("") and then decode it here: https://www.base64decode.org/ – Nathan B Mar 14 '21 at 15:39
159

Internet Explorer 10+

// Define the string
var string = 'Hello World!';

// Encode the String
var encodedString = btoa(string);
console.log(encodedString); // Outputs: "SGVsbG8gV29ybGQh"

// Decode the String
var decodedString = atob(encodedString);
console.log(decodedString); // Outputs: "Hello World!"

Cross-Browser

// Create Base64 Object
var Base64={_keyStr:"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=",encode:function(e){var t="";var n,r,i,s,o,u,a;var f=0;e=Base64._utf8_encode(e);while(f<e.length){n=e.charCodeAt(f++);r=e.charCodeAt(f++);i=e.charCodeAt(f++);s=n>>2;o=(n&3)<<4|r>>4;u=(r&15)<<2|i>>6;a=i&63;if(isNaN(r)){u=a=64}else if(isNaN(i)){a=64}t=t+this._keyStr.charAt(s)+this._keyStr.charAt(o)+this._keyStr.charAt(u)+this._keyStr.charAt(a)}return t},decode:function(e){var t="";var n,r,i;var s,o,u,a;var f=0;e=e.replace(/[^A-Za-z0-9\+\/\=]/g,"");while(f<e.length){s=this._keyStr.indexOf(e.charAt(f++));o=this._keyStr.indexOf(e.charAt(f++));u=this._keyStr.indexOf(e.charAt(f++));a=this._keyStr.indexOf(e.charAt(f++));n=s<<2|o>>4;r=(o&15)<<4|u>>2;i=(u&3)<<6|a;t=t+String.fromCharCode(n);if(u!=64){t=t+String.fromCharCode(r)}if(a!=64){t=t+String.fromCharCode(i)}}t=Base64._utf8_decode(t);return t},_utf8_encode:function(e){e=e.replace(/\r\n/g,"\n");var t="";for(var n=0;n<e.length;n++){var r=e.charCodeAt(n);if(r<128){t+=String.fromCharCode(r)}else if(r>127&&r<2048){t+=String.fromCharCode(r>>6|192);t+=String.fromCharCode(r&63|128)}else{t+=String.fromCharCode(r>>12|224);t+=String.fromCharCode(r>>6&63|128);t+=String.fromCharCode(r&63|128)}}return t},_utf8_decode:function(e){var t="";var n=0;var r=c1=c2=0;while(n<e.length){r=e.charCodeAt(n);if(r<128){t+=String.fromCharCode(r);n++}else if(r>191&&r<224){c2=e.charCodeAt(n+1);t+=String.fromCharCode((r&31)<<6|c2&63);n+=2}else{c2=e.charCodeAt(n+1);c3=e.charCodeAt(n+2);t+=String.fromCharCode((r&15)<<12|(c2&63)<<6|c3&63);n+=3}}return t}}

// Define the string
var string = 'Hello World!';

// Encode the String
var encodedString = Base64.encode(string);
console.log(encodedString); // Outputs: "SGVsbG8gV29ybGQh"

// Decode the String
var decodedString = Base64.decode(encodedString);
console.log(decodedString); // Outputs: "Hello World!"

jsFiddle


with Node.js

Here is how you encode normal text to base64 in Node.js:

//Buffer() requires a number, array or string as the first parameter, and an optional encoding type as the second parameter. 
// Default is utf8, possible encoding types are ascii, utf8, ucs2, base64, binary, and hex
var b = new Buffer('JavaScript');
// If we don't use toString(), JavaScript assumes we want to convert the object to utf8.
// We can make it convert to other formats by passing the encoding type to toString().
var s = b.toString('base64');

And here is how you decode base64 encoded strings:

var b = new Buffer('SmF2YVNjcmlwdA==', 'base64')
var s = b.toString();

with Dojo.js

To encode an array of bytes using dojox.encoding.base64:

var str = dojox.encoding.base64.encode(myByteArray);

To decode a base64-encoded string:

var bytes = dojox.encoding.base64.decode(str)

bower install angular-base64

<script src="bower_components/angular-base64/angular-base64.js"></script>

angular
    .module('myApp', ['base64'])
    .controller('myController', [

    '$base64', '$scope', 
    function($base64, $scope) {

        $scope.encoded = $base64.encode('a string');
        $scope.decoded = $base64.decode('YSBzdHJpbmc=');
}]);
davidcondrey
  • 29,530
  • 14
  • 105
  • 129
96

Sunny's code is great except it breaks in IE7 because of references to "this". Fixed by replacing such references with "Base64":

var Base64 = {
// private property
_keyStr : "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=",

// public method for encoding
encode : function (input) {
    var output = "";
    var chr1, chr2, chr3, enc1, enc2, enc3, enc4;
    var i = 0;

    input = Base64._utf8_encode(input);

    while (i < input.length) {

        chr1 = input.charCodeAt(i++);
        chr2 = input.charCodeAt(i++);
        chr3 = input.charCodeAt(i++);

        enc1 = chr1 >> 2;
        enc2 = ((chr1 & 3) << 4) | (chr2 >> 4);
        enc3 = ((chr2 & 15) << 2) | (chr3 >> 6);
        enc4 = chr3 & 63;

        if (isNaN(chr2)) {
            enc3 = enc4 = 64;
        } else if (isNaN(chr3)) {
            enc4 = 64;
        }

        output = output +
        Base64._keyStr.charAt(enc1) + Base64._keyStr.charAt(enc2) +
        Base64._keyStr.charAt(enc3) + Base64._keyStr.charAt(enc4);

    }

    return output;
},

// public method for decoding
decode : function (input) {
    var output = "";
    var chr1, chr2, chr3;
    var enc1, enc2, enc3, enc4;
    var i = 0;

    input = input.replace(/[^A-Za-z0-9\+\/\=]/g, "");

    while (i < input.length) {

        enc1 = Base64._keyStr.indexOf(input.charAt(i++));
        enc2 = Base64._keyStr.indexOf(input.charAt(i++));
        enc3 = Base64._keyStr.indexOf(input.charAt(i++));
        enc4 = Base64._keyStr.indexOf(input.charAt(i++));

        chr1 = (enc1 << 2) | (enc2 >> 4);
        chr2 = ((enc2 & 15) << 4) | (enc3 >> 2);
        chr3 = ((enc3 & 3) << 6) | enc4;

        output = output + String.fromCharCode(chr1);

        if (enc3 != 64) {
            output = output + String.fromCharCode(chr2);
        }
        if (enc4 != 64) {
            output = output + String.fromCharCode(chr3);
        }

    }

    output = Base64._utf8_decode(output);

    return output;

},

// private method for UTF-8 encoding
_utf8_encode : function (string) {
    string = string.replace(/\r\n/g,"\n");
    var utftext = "";

    for (var n = 0; n < string.length; n++) {

        var c = string.charCodeAt(n);

        if (c < 128) {
            utftext += String.fromCharCode(c);
        }
        else if((c > 127) && (c < 2048)) {
            utftext += String.fromCharCode((c >> 6) | 192);
            utftext += String.fromCharCode((c & 63) | 128);
        }
        else {
            utftext += String.fromCharCode((c >> 12) | 224);
            utftext += String.fromCharCode(((c >> 6) & 63) | 128);
            utftext += String.fromCharCode((c & 63) | 128);
        }

    }

    return utftext;
},

// private method for UTF-8 decoding
_utf8_decode : function (utftext) {
    var string = "";
    var i = 0;
    var c = c1 = c2 = 0;

    while ( i < utftext.length ) {

        c = utftext.charCodeAt(i);

        if (c < 128) {
            string += String.fromCharCode(c);
            i++;
        }
        else if((c > 191) && (c < 224)) {
            c2 = utftext.charCodeAt(i+1);
            string += String.fromCharCode(((c & 31) << 6) | (c2 & 63));
            i += 2;
        }
        else {
            c2 = utftext.charCodeAt(i+1);
            c3 = utftext.charCodeAt(i+2);
            string += String.fromCharCode(((c & 15) << 12) | ((c2 & 63) << 6) | (c3 & 63));
            i += 3;
        }

    }
    return string;
}
}
user850789
  • 961
  • 6
  • 2
  • 4
    ooh my bad, I was taking the input from browser URL; where | is converted to %7C; hence the encoding also wrong. – Kanagavelu Sugumar Sep 11 '13 at 08:03
  • I know this is really old, but I have seen this function used in more than one place, the key string is actually at 65 characters, not 64. The string is not standard spec, I am not sure it matters, but was just wondering if it does? – Jonathan Wagner Jun 17 '15 at 04:17
  • "use strict"; is what breaks the 'this' and other type elements like 'with' and from what I have read, 'eval' gets a bashing. All misplaced ideas on abuse. Personally I do not see why JavaScript needs to go down the route its going, it was never meant to be a program tightly bound and made more complex than it is already. If you want to be bound then make a compiler for javascript. – Mark Giblin Oct 10 '15 at 12:30
  • I try to use this function and I receive the error: Caused by: org.mozilla.javascript.EcmaError: TypeError: Cannot find function replace in object teste teste teste I'm trying to encode .txt with "teste teste teste". Anyone knows why this error? – PRVS Nov 04 '15 at 09:19
  • @JonathanWagner - there are 64 characters used for the normal encoding. The 65th character is used as padding them the input string does not have a number of characters divisible by 3. – Kickstart Sep 05 '17 at 13:58
93

You can use btoa (to base-64) and atob (from base-64).

For IE 9 and below, try the jquery-base64 plugin:

$.base64.encode("this is a test");
$.base64.decode("dGhpcyBpcyBhIHRlc3Q=");
Vitalii Fedorenko
  • 97,155
  • 26
  • 144
  • 111
  • 139
    Why does everything need to be a jQuery plugin :c this is just core JavaScript functionality this has nothing to do with the DOM or jQuery – EaterOfCode Apr 29 '13 at 11:04
  • 39
    This is not a core functionality or there wouldn't be as many different high voted answers (including do-it-yourself tl;dr code). So, imho this is actually a good use case for jQuery (one liner, expected to work even in Android's WebView) - even more if it's already a dependency. – Risadinha Aug 26 '13 at 17:04
  • I've gone and given up. A jQuery subset should be formalized and introduced into the JS standard like a ".net". It's not pretty, but the need is overwhelming and it's come to rule the ecosystem. – Lodewijk Jun 14 '14 at 16:52
  • 1
    I like to install code snippets like this into jQuery mainly because they'll exist in a controlled namespace. If you're not using AMD or CommonJS or a similar design pattern, it's easy for your global namespace to get really messy with a bunch of random functions. – sffc Jun 25 '14 at 07:23
  • http://www.w3schools.com/jsref/met_win_btoa.asp indicates btoa support for chrome, ie10+, firefox, safari, opera – OzBob Jul 22 '15 at 02:11
  • This "Global Namespace" business, and using frameworks for execution of small programs, bag idea and JQuery is written in JavaScript, IT IS NOT JavaScript. Read up about atob() and btoa() because you will find that it is not consistent across browser types. – Mark Giblin Oct 10 '15 at 12:32
  • 10
    @Risadinha - except its functionality does not depend on or extend anything jQuery at all...literally the only references to jQuery in its code are attaching it to the jQuery object...so what's the point in attaching it to jQuery and therefore requiring jQuery to use? Just make it it's own 1 liner `base64.encode(...)` and `base64.decode(...)` ...attaching it to jQuery when it has zero jQuery specific functionality makes absolutely no sense... – Jimbo Jonny Mar 12 '16 at 05:28
  • 1
    jQuery was not requested. Not a valid answer to a plain old JS question. – metaColin Feb 10 '17 at 22:20
46

From the comments (by SET and Stefan Steiger) below the accepted answer, here is a quick summary of how to encode/decode a string to/from base64 without need of a library.

str = "The quick brown fox jumps over the lazy dog";
b64 = btoa(unescape(encodeURIComponent(str)));
str = decodeURIComponent(escape(window.atob(b64)));

Demo

(uses jQuery library, but not for encode/decode)

str = "The quick brown fox jumps over the lazy dog";

$('input').val(str);

$('#btnConv').click(function(){
  var txt = $('input').val();
  var b64 = btoa(unescape(encodeURIComponent(txt)));
  $('input').val(b64);
  $('#btnDeConv').show();
});
$('#btnDeConv').click(function(){
  var b64 = $('input').val();
  var txt = decodeURIComponent(escape(window.atob(b64)));
  $('input').val(txt);
});
#btnDeConv{display:none;}
<script src="https://ajax.googleapis.com/ajax/libs/jquery/2.1.1/jquery.min.js"></script>

<input type="text" />
<button id="btnConv">Convert</button>
<button id="btnDeConv">DeConvert</button>
Crashalot
  • 31,452
  • 56
  • 235
  • 393
crashwap
  • 2,324
  • 3
  • 24
  • 44
  • To confirm, this supports UTF-8 characters? – Crashalot Feb 03 '18 at 09:09
  • 1
    @Crashalot I realize this is two years too late, but yes it does. I am also just realizing as I type this that you provided an edit that possibly made UTF8 work. – tycrek Apr 05 '20 at 17:34
  • For anyone here looking for a good solution to use with Node.js, I can confirm this works. For decoding in Node, I used: `Buffer.from(b64data, 'base64').toString();` – tycrek Apr 05 '20 at 17:35
  • Can someone explain what `unescape` and `escape` are doing in this solution snipped? – Sammi Sep 16 '20 at 16:09
27

There's a couple of bugs in both implementations of _utf8_decode. c1 and c2 are assigned as global variables due to broken use of the var statement, and c3 is not initialized or declared at all.

It works, but these variables will overwrite any existing ones with the same name outside this function.

Here's a version that won't do this:

// private method for UTF-8 decoding
_utf8_decode : function (utftext) {
    var string = "";
    var i = 0;
    var c = 0, c1 = 0, c2 = 0;

    while ( i < utftext.length ) {

        c = utftext.charCodeAt(i);

        if (c < 128) {
            string += String.fromCharCode(c);
            i++;
        }
        else if((c > 191) && (c < 224)) {
            c1 = utftext.charCodeAt(i+1);
            string += String.fromCharCode(((c & 31) << 6) | (c1 & 63));
            i += 2;
        }
        else {
            c1 = utftext.charCodeAt(i+1);
            c2 = utftext.charCodeAt(i+2);
            string += String.fromCharCode(((c & 15) << 12) | ((c1 & 63) << 6) | (c2 & 63));
            i += 3;
        }

    }
    return string;
}
robbles
  • 2,421
  • 1
  • 21
  • 29
  • 9
    @Daan I didn't have enough rep to edit answers when I wrote this answer...in 2011. – robbles Dec 10 '13 at 17:42
  • 2
    IE7 ? i guess we should stop wasting time to write code for that, people won't stop using this old technology unless we developers forced them to! – Rami Dabain Jun 07 '14 at 15:01
  • @RonanDejhero does it not work in IE7? I don't remember if I tested in that particular browser. – robbles Jun 08 '14 at 22:47
  • 1
    What i meant that if it does not work in IE7, no one should care!. i didn't test and won't test it :) – Rami Dabain Jun 09 '14 at 11:02
16

I +1'ed Sunny's answer, but I wanted to contribute back a few changes I made for my own project in case anyone should find it useful. Basically I've just cleaned up the original code a little so JSLint doesn't complain quite as much, and I made the methods marked as private in the comments actually private. I also added two methods I needed in my own project, namely decodeToHex and encodeFromHex.

The code:

var Base64 = (function() {
    "use strict";

    var _keyStr = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=";

    var _utf8_encode = function (string) {

        var utftext = "", c, n;

        string = string.replace(/\r\n/g,"\n");

        for (n = 0; n < string.length; n++) {

            c = string.charCodeAt(n);

            if (c < 128) {

                utftext += String.fromCharCode(c);

            } else if((c > 127) && (c < 2048)) {

                utftext += String.fromCharCode((c >> 6) | 192);
                utftext += String.fromCharCode((c & 63) | 128);

            } else {

                utftext += String.fromCharCode((c >> 12) | 224);
                utftext += String.fromCharCode(((c >> 6) & 63) | 128);
                utftext += String.fromCharCode((c & 63) | 128);

            }

        }

        return utftext;
    };

    var _utf8_decode = function (utftext) {
        var string = "", i = 0, c = 0, c1 = 0, c2 = 0;

        while ( i < utftext.length ) {

            c = utftext.charCodeAt(i);

            if (c < 128) {

                string += String.fromCharCode(c);
                i++;

            } else if((c > 191) && (c < 224)) {

                c1 = utftext.charCodeAt(i+1);
                string += String.fromCharCode(((c & 31) << 6) | (c1 & 63));
                i += 2;

            } else {

                c1 = utftext.charCodeAt(i+1);
                c2 = utftext.charCodeAt(i+2);
                string += String.fromCharCode(((c & 15) << 12) | ((c1 & 63) << 6) | (c2 & 63));
                i += 3;

            }

        }

        return string;
    };

    var _hexEncode = function(input) {
        var output = '', i;

        for(i = 0; i < input.length; i++) {
            output += input.charCodeAt(i).toString(16);
        }

        return output;
    };

    var _hexDecode = function(input) {
        var output = '', i;

        if(input.length % 2 > 0) {
            input = '0' + input;
        }

        for(i = 0; i < input.length; i = i + 2) {
            output += String.fromCharCode(parseInt(input.charAt(i) + input.charAt(i + 1), 16));
        }

        return output;
    };

    var encode = function (input) {
        var output = "", chr1, chr2, chr3, enc1, enc2, enc3, enc4, i = 0;

        input = _utf8_encode(input);

        while (i < input.length) {

            chr1 = input.charCodeAt(i++);
            chr2 = input.charCodeAt(i++);
            chr3 = input.charCodeAt(i++);

            enc1 = chr1 >> 2;
            enc2 = ((chr1 & 3) << 4) | (chr2 >> 4);
            enc3 = ((chr2 & 15) << 2) | (chr3 >> 6);
            enc4 = chr3 & 63;

            if (isNaN(chr2)) {
                enc3 = enc4 = 64;
            } else if (isNaN(chr3)) {
                enc4 = 64;
            }

            output += _keyStr.charAt(enc1);
            output += _keyStr.charAt(enc2);
            output += _keyStr.charAt(enc3);
            output += _keyStr.charAt(enc4);

        }

        return output;
    };

    var decode = function (input) {
        var output = "", chr1, chr2, chr3, enc1, enc2, enc3, enc4, i = 0;

        input = input.replace(/[^A-Za-z0-9\+\/\=]/g, "");

        while (i < input.length) {

            enc1 = _keyStr.indexOf(input.charAt(i++));
            enc2 = _keyStr.indexOf(input.charAt(i++));
            enc3 = _keyStr.indexOf(input.charAt(i++));
            enc4 = _keyStr.indexOf(input.charAt(i++));

            chr1 = (enc1 << 2) | (enc2 >> 4);
            chr2 = ((enc2 & 15) << 4) | (enc3 >> 2);
            chr3 = ((enc3 & 3) << 6) | enc4;

            output += String.fromCharCode(chr1);

            if (enc3 !== 64) {
                output += String.fromCharCode(chr2);
            }
            if (enc4 !== 64) {
                output += String.fromCharCode(chr3);
            }

        }

        return _utf8_decode(output);
    };

    var decodeToHex = function(input) {
        return _hexEncode(decode(input));
    };

    var encodeFromHex = function(input) {
        return encode(_hexDecode(input));
    };

    return {
        'encode': encode,
        'decode': decode,
        'decodeToHex': decodeToHex,
        'encodeFromHex': encodeFromHex
    };
}());
Joe Dyndale
  • 1,075
  • 1
  • 14
  • 31
  • I initially thought your unrolling of the output concatenation into separate statements would be more optimal, but after I thought about it for a sec, this *should* be more inefficient since Javascript strings are immutable and it would cause 4 copies of potentially huge data blobs when working with large binary data files. It is a safer bet to concatenate the 4 chars together first and then building a new string. I wish I knew for certain of a better string building method that would be sure to be efficient on all platforms. (even IE6) – Marius Jan 04 '13 at 18:19
  • I haven't considered performance in my cleanup of the originally posted code. I just made it more readable and made methods marked as private in the comments in the original actually be private by using the revealing module pattern. I'm sure it can be optimized in regards to performance as well. Not quite sure when garbage collection would kick in here, and hashing large files through Javascript isn't very common (or indeed likely not the optimal solution in any case). – Joe Dyndale Jan 07 '13 at 11:39
  • Funny how this code sort of lives here. There are already 3 different versions of it on this page. – gregn3 Mar 07 '14 at 12:37
16

This question and it's answers pointed me to the right direction.
Especially with unicode atob and btoa can not be used "vanilla" and these days EVERYTHING is unicode ..

Directly from Mozilla, two nice functions for this purpose (tested with unicode and html tags inside)

function b64EncodeUnicode(str) {
    return btoa(encodeURIComponent(str).replace(/%([0-9A-F]{2})/g, function(match, p1) {
        return String.fromCharCode('0x' + p1);
    }));
}

b64EncodeUnicode('✓ à la mode'); // "4pyTIMOgIGxhIG1vZGU="
b64EncodeUnicode('\n'); // "Cg=="



function b64DecodeUnicode(str) {
    return decodeURIComponent(Array.prototype.map.call(atob(str), function(c) {
        return '%' + ('00' + c.charCodeAt(0).toString(16)).slice(-2);
    }).join(''));
}

b64DecodeUnicode('4pyTIMOgIGxhIG1vZGU='); // "✓ à la mode"
b64DecodeUnicode('Cg=='); // "\n"

These functions will perform lightning fast in comparison to raw base64 decoding using a custom javascript function as btoa and atob are executed outside the interpreter.

If you can ignore old IE and old mobile phones (like iphone 3?) this should be a good solution.

John
  • 6,400
  • 2
  • 44
  • 49
16

For newer browsers to encode Uint8Array to string, and decode string to Uint8Array.

const base64 = {
    decode: s => Uint8Array.from(atob(s), c => c.charCodeAt(0)),
    encode: b => btoa(String.fromCharCode(...new Uint8Array(b)))
};

For Node.js you can use the following to encode string, Buffer, or Uint8Array to string, and decode from string, Buffer, or Uint8Array to Buffer.

const base64 = {
    decode: s => Buffer.from(s, 'base64'),
    encode: b => Buffer.from(b).toString('base64')
};
Rix
  • 1,466
  • 12
  • 19
13

Please note that this is not suitable for raw Unicode strings! See Unicode section here.

Syntax for encoding

var encodedData = window.btoa(stringToEncode);

Syntax for decoding

var decodedData = window.atob(encodedData);

Kathir
  • 1,090
  • 14
  • 22
  • 1
    Direct link to the unicode section: https://developer.mozilla.org/en-US/docs/Web/API/WindowOrWorkerGlobalScope/btoa#Unicode_strings – TomTasche Jan 25 '20 at 10:17
13

To make a Base64 encoded String URL friendly, in JavaScript you could do something like this:

// if this is your Base64 encoded string
var str = 'VGhpcyBpcyBhbiBhd2Vzb21lIHNjcmlwdA=='; 

// make URL friendly:
str = str.replace(/\+/g, '-').replace(/\//g, '_').replace(/\=+$/, '');

// reverse to original encoding
str = (str + '===').slice(0, str.length + (str.length % 4));
str = str.replace(/-/g, '+').replace(/_/g, '/');

See also this Fiddle: http://jsfiddle.net/magikMaker/7bjaT/

magikMaker
  • 2,654
  • 25
  • 20
  • 10
    I would humbly suggest that the use of `encodeURIComponent` may well result in a superior outcome with less expenditure of effort on the part of the developer. – Pablo Fernandez Oct 28 '11 at 16:22
  • 11
    encodeURIComponent will change the length of base64 encoded strings, and replacing '-' and '_' with '+' and '/' is standard practice when using base64 in URLs (e.g. http://docs.python.org/library/base64.html#base64.urlsafe_b64encode). No need to get upset. – natevw Dec 22 '11 at 17:29
12

I have re-wrote by hand, these encoding and decoding methods with the exception of the hexadecimal one into a modular format for cross-platform / browser compatibility and also with real private scoping, and uses btoa and atob if they exist due to speed rather than utilize its own encoding:

https://gist.github.com/Nijikokun/5192472

Usage:

base64.encode(/* String */);
base64.decode(/* String */);

utf8.encode(/* String */);
utf8.decode(/* String */);
Nijikokun
  • 1,396
  • 1
  • 14
  • 22
10

if you need to encode HTML image object, you can write simple function like:

function getBase64Image(img) {  
  var canvas = document.createElement("canvas");  
  canvas.width = img.width;  
  canvas.height = img.height;  
  var ctx = canvas.getContext("2d");  
  ctx.drawImage(img, 0, 0);  
  var dataURL = canvas.toDataURL("image/png");  
  // escape data:image prefix
  return dataURL.replace(/^data:image\/(png|jpg);base64,/, "");  
  // or just return dataURL
  // return dataURL
}  

To get base64 of image by id:

function getBase64ImageById(id){  
  return getBase64Image(document.getElementById(id));  
} 

more here

Nedudi
  • 4,814
  • 2
  • 35
  • 31
8

I needed encoding of an UTF-8 string as base64 for a project of mine. Most of the answers here don't seem to properly handle UTF-16 surrogate pairs when converting to UTF-8 so, for completion sake, I will post my solution:

function strToUTF8Base64(str) {

    function decodeSurrogatePair(hi, lo) {
        var resultChar = 0x010000;
        resultChar += lo - 0xDC00;
        resultChar += (hi - 0xD800) << 10;
        return resultChar;
    }

    var bytes = [0, 0, 0];
    var byteIndex = 0;
    var result = [];

    function output(s) {
        result.push(s);
    }

    function emitBase64() {

        var digits =
                'ABCDEFGHIJKLMNOPQRSTUVWXYZ' +
                'abcdefghijklmnopqrstuvwxyz' +
                '0123456789+/';

        function toDigit(value) {
            return digits[value];
        }

        // --Byte 0--    --Byte 1--    --Byte 2--
        // 1111  1122    2222  3333    3344  4444

        var d1 = toDigit(bytes[0] >> 2);
        var d2 = toDigit(
            ((bytes[0] & 0x03) << 4) |
            (bytes[1] >> 4));
        var d3 = toDigit(
            ((bytes[1] & 0x0F) << 2) |
            (bytes[2] >> 6));
        var d4 = toDigit(
            bytes[2] & 0x3F);

        if (byteIndex === 1) {
            output(d1 + d2 + '==');
        }
        else if (byteIndex === 2) {
            output(d1 + d2 + d3 + '=');
        }
        else {
            output(d1 + d2 + d3 + d4);
        }
    }

    function emit(chr) {
        bytes[byteIndex++] = chr;
        if (byteIndex == 3) {
            emitBase64();
            bytes[0] = 0;
            bytes[1] = 0;
            bytes[2] = 0;
            byteIndex = 0;
        }
    }

    function emitLast() {
        if (byteIndex > 0) {
            emitBase64();
        }
    }

    // Converts the string to UTF8:

    var i, chr;
    var hi, lo;
    for (i = 0; i < str.length; i++) {
        chr = str.charCodeAt(i);

        // Test and decode surrogate pairs in the string
        if (chr >= 0xD800 && chr <= 0xDBFF) {
            hi = chr;
            lo = str.charCodeAt(i + 1);
            if (lo >= 0xDC00 && lo <= 0xDFFF) {
                chr = decodeSurrogatePair(hi, lo);
                i++;
            }
        }

        // Encode the character as UTF-8.
        if (chr < 0x80) {
            emit(chr);
        }
        else if (chr < 0x0800) {
            emit((chr >> 6) | 0xC0);
            emit(((chr >> 0) & 0x3F) | 0x80);
        }
        else if (chr < 0x10000) {
            emit((chr >> 12) | 0xE0);
            emit(((chr >>  6) & 0x3F) | 0x80);
            emit(((chr >>  0) & 0x3F) | 0x80);
        }
        else if (chr < 0x110000) {
            emit((chr >> 18) | 0xF0);
            emit(((chr >> 12) & 0x3F) | 0x80);
            emit(((chr >>  6) & 0x3F) | 0x80);
            emit(((chr >>  0) & 0x3F) | 0x80);
        }
    }

    emitLast();

    return result.join('');
}

Note that the code is not thoroughly tested. I tested some inputs, including things like strToUTF8Base64('衠衢蠩蠨') and compared with the output of an online encoding tool (https://www.base64encode.org/).

Ricardo Machado
  • 309
  • 3
  • 4
  • This is a great job! But instead of "return digits[value]" it should be "return digits.substr(value,1)" in function emitBase64().becose error/ Stumbles over emoji and multibyte characters. But it often encodes what online decoders cannot decode! This code brought out the basic Cyrillic alphabet without problems. – Garric Apr 28 '21 at 00:24
7

Contributing with a minified polyfill for window.atob + window.btoa that I'm currently using.

(function(){function t(t){this.message=t}var e="undefined"!=typeof exports?exports:this,r="ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=";t.prototype=Error(),t.prototype.name="InvalidCharacterError",e.btoa||(e.btoa=function(e){for(var o,n,a=0,i=r,c="";e.charAt(0|a)||(i="=",a%1);c+=i.charAt(63&o>>8-8*(a%1))){if(n=e.charCodeAt(a+=.75),n>255)throw new t("'btoa' failed: The string to be encoded contains characters outside of the Latin1 range.");o=o<<8|n}return c}),e.atob||(e.atob=function(e){if(e=e.replace(/=+$/,""),1==e.length%4)throw new t("'atob' failed: The string to be decoded is not correctly encoded.");for(var o,n,a=0,i=0,c="";n=e.charAt(i++);~n&&(o=a%4?64*o+n:n,a++%4)?c+=String.fromCharCode(255&o>>(6&-2*a)):0)n=r.indexOf(n);return c})})();
Johan
  • 31,613
  • 48
  • 166
  • 272
6

I'd rather use the bas64 encode/decode methods from CryptoJS, the most popular library for standard and secure cryptographic algorithms implemented in JavaScript using best practices and patterns.

Dan Dascalescu
  • 110,650
  • 40
  • 276
  • 363
6

Here is an AngularJS Factory version of @user850789's one:

'use strict';

var ProjectNameBase64Factory = angular.module('project_name.factories.base64', []);

ProjectNameBase64Factory.factory('Base64', function () {
    var Base64 = {
        // private property
        _keyStr: "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=",

        // public method for encoding
        encode: function (input) {
            var output = "";
            var chr1, chr2, chr3, enc1, enc2, enc3, enc4;
            var i = 0;

            input = Base64._utf8_encode(input);

            while (i < input.length) {

                chr1 = input.charCodeAt(i++);
                chr2 = input.charCodeAt(i++);
                chr3 = input.charCodeAt(i++);

                enc1 = chr1 >> 2;
                enc2 = ((chr1 & 3) << 4) | (chr2 >> 4);
                enc3 = ((chr2 & 15) << 2) | (chr3 >> 6);
                enc4 = chr3 & 63;

                if (isNaN(chr2)) {
                    enc3 = enc4 = 64;
                } else if (isNaN(chr3)) {
                    enc4 = 64;
                }

                output = output +
                         Base64._keyStr.charAt(enc1) + Base64._keyStr.charAt(enc2) +
                         Base64._keyStr.charAt(enc3) + Base64._keyStr.charAt(enc4);

            }

            return output;
        },

        // public method for decoding
        decode: function (input) {
            var output = "";
            var chr1, chr2, chr3;
            var enc1, enc2, enc3, enc4;
            var i = 0;

            input = input.replace(/[^A-Za-z0-9\+\/\=]/g, "");

            while (i < input.length) {

                enc1 = Base64._keyStr.indexOf(input.charAt(i++));
                enc2 = Base64._keyStr.indexOf(input.charAt(i++));
                enc3 = Base64._keyStr.indexOf(input.charAt(i++));
                enc4 = Base64._keyStr.indexOf(input.charAt(i++));

                chr1 = (enc1 << 2) | (enc2 >> 4);
                chr2 = ((enc2 & 15) << 4) | (enc3 >> 2);
                chr3 = ((enc3 & 3) << 6) | enc4;

                output = output + String.fromCharCode(chr1);

                if (enc3 != 64) {
                    output = output + String.fromCharCode(chr2);
                }
                if (enc4 != 64) {
                    output = output + String.fromCharCode(chr3);
                }

            }

            output = Base64._utf8_decode(output);

            return output;

        },

        // private method for UTF-8 encoding
        _utf8_encode: function (string) {
            string = string.replace(/\r\n/g, "\n");
            var utftext = "";

            for (var n = 0; n < string.length; n++) {

                var c = string.charCodeAt(n);

                if (c < 128) {
                    utftext += String.fromCharCode(c);
                }
                else if ((c > 127) && (c < 2048)) {
                    utftext += String.fromCharCode((c >> 6) | 192);
                    utftext += String.fromCharCode((c & 63) | 128);
                }
                else {
                    utftext += String.fromCharCode((c >> 12) | 224);
                    utftext += String.fromCharCode(((c >> 6) & 63) | 128);
                    utftext += String.fromCharCode((c & 63) | 128);
                }

            }

            return utftext;
        },

        // private method for UTF-8 decoding
        _utf8_decode: function (utftext) {
            var string = "";
            var i = 0;
            var c = 0, c2 = 0, c3 = 0;

            while (i < utftext.length) {

                c = utftext.charCodeAt(i);

                if (c < 128) {
                    string += String.fromCharCode(c);
                    i++;
                }
                else if ((c > 191) && (c < 224)) {
                    c2 = utftext.charCodeAt(i + 1);
                    string += String.fromCharCode(((c & 31) << 6) | (c2 & 63));
                    i += 2;
                }
                else {
                    c2 = utftext.charCodeAt(i + 1);
                    c3 = utftext.charCodeAt(i + 2);
                    string += String.fromCharCode(((c & 15) << 12) | ((c2 & 63) << 6) | (c3 & 63));
                    i += 3;
                }

            }
            return string;
        }
    };
    return Base64;
});
A T
  • 10,508
  • 14
  • 85
  • 137
5

For my project I still need to support IE7 and work with large input to encode.

Based on the code proposed by Joe Dyndale and as suggested in comment by Marius, it is possible to improve the performance with IE7 by constructing the result with an array instead of a string.

Here is the example for encode:

var encode = function (input) {
    var output = [], chr1, chr2, chr3, enc1, enc2, enc3, enc4, i = 0;

    input = _utf8_encode(input);

    while (i < input.length) {

        chr1 = input.charCodeAt(i++);
        chr2 = input.charCodeAt(i++);
        chr3 = input.charCodeAt(i++);

        enc1 = chr1 >> 2;
        enc2 = ((chr1 & 3) << 4) | (chr2 >> 4);
        enc3 = ((chr2 & 15) << 2) | (chr3 >> 6);
        enc4 = chr3 & 63;

        if (isNaN(chr2)) {
            enc3 = enc4 = 64;
        } else if (isNaN(chr3)) {
            enc4 = 64;
        }

        output.push(_keyStr.charAt(enc1));
        output.push(_keyStr.charAt(enc2));
        output.push(_keyStr.charAt(enc3));
        output.push(_keyStr.charAt(enc4));

    }

    return output.join("");
};
Nico
  • 59
  • 1
  • 2
5

Well, if you are using dojo, it gives us direct way to encode or decode into base64.

Try this:-

To encode an array of bytes using dojox.encoding.base64:

var str = dojox.encoding.base64.encode(myByteArray);

To decode a base64-encoded string:

var bytes = dojox.encoding.base64.decode(str);
Camilo Terevinto
  • 26,697
  • 6
  • 67
  • 99
Vikash Pandey
  • 5,073
  • 6
  • 36
  • 41
5

While a bit more work, if you want a high performance native solution there are some HTML5 functions you can use.

If you can get your data into a Blob, then you can use the FileReader.readAsDataURL() function to get a data:// URL and chop off the front of it to get at the base64 data.

You may have to do further processing however to urldecode the data, as I'm not sure whether + characters are escaped or not for the data:// URL, but this should be pretty trivial.

Malvineous
  • 20,120
  • 14
  • 99
  • 121
5

You can use window.btoa and window.atob...

const encoded = window.btoa('Alireza Dezfoolian'); // encode a string
const decoded = window.atob(encoded); // decode the string

Probably using the way which MDN is can do your job the best... Also accepting unicode... using these two simple functions:

// ucs-2 string to base64 encoded ascii
function utoa(str) {
    return window.btoa(unescape(encodeURIComponent(str)));
}
// base64 encoded ascii to ucs-2 string
function atou(str) {
    return decodeURIComponent(escape(window.atob(str)));
}
// Usage:
utoa('✓ à la mode'); // 4pyTIMOgIGxhIG1vZGU=
atou('4pyTIMOgIGxhIG1vZGU='); // "✓ à la mode"

utoa('I \u2661 Unicode!'); // SSDimaEgVW5pY29kZSE=
atou('SSDimaEgVW5pY29kZSE='); // "I ♡ Unicode!"
Alireza
  • 83,698
  • 19
  • 241
  • 152
5

Here is a LIVE DEMO of atob() and btoa() JS built in functions:

<!DOCTYPE html>
<html>
  <head>
    <style>
      textarea{
        width:30%;
        height:100px;
      }
    </style>
    <script>
      // encode string to base64
      function encode()
      {
        var txt = document.getElementById("txt1").value;
        var result = btoa(txt);
        document.getElementById("txt2").value = result;
      }
      // decode base64 back to original string
      function decode()
      {
        var txt = document.getElementById("txt3").value;
        var result = atob(txt);
        document.getElementById("txt4").value = result;
      }
    </script>
  </head>
  <body>
    <div>
      <textarea id="txt1">Some text to decode
      </textarea>
    </div>
    <div>
      <input type="button" id="btnencode" value="Encode" onClick="encode()"/>
    </div>
    <div>
      <textarea id="txt2">
      </textarea>
    </div>
    <br/>
    <div>
      <textarea id="txt3">U29tZSB0ZXh0IHRvIGRlY29kZQ==
      </textarea>
    </div>
    <div>
      <input type="button" id="btndecode" value="Decode" onClick="decode()"/>
    </div>
    <div>
      <textarea id="txt4">
      </textarea>
    </div>
  </body>
</html>
jonathana
  • 4,633
  • 3
  • 25
  • 42
3
let a = Buffer.from('JavaScript').toString('base64');
console.log(a);

let b = Buffer.from(a, 'base64').toString();
console.log(b);
qux
  • 119
  • 6
2

Use js-base64 library as

btoa() doesn't work with emojis

var str = "I was funny ";
console.log("Original string:", str);

var encodedStr = Base64.encode(str)
console.log("Encoded string:", encodedStr);

var decodedStr = Base64.decode(encodedStr)
console.log("Decoded string:", decodedStr);
<script src="https://cdn.jsdelivr.net/npm/js-base64@2.5.2/base64.min.js"></script>
Shivaji Mutkule
  • 385
  • 1
  • 6
  • 19
1

You can use btoa()/atob() in browser, but some improvements required, as described here https://base64tool.com/uncaught-domexception-btoa-on-window/ and there https://developer.mozilla.org/en-US/docs/Web/API/WindowOrWorkerGlobalScope/btoa for UTF strings support!

Vladimir Ignatyev
  • 1,783
  • 15
  • 30
1

JS without btoa middlestep (no lib)

In question title you write about string conversion, but in question you talk about binary data (picture) so here is function which make proper conversion starting from PNG picture binary data (details and reversal conversion here )

enter image description here

function bytesArrToBase64(arr) {
  const abc = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"; // base64 alphabet
  const bin = n => n.toString(2).padStart(8,0); // convert num to 8-bit binary string
  const l = arr.length
  let result = '';

  for(let i=0; i<=(l-1)/3; i++) {
    let c1 = i*3+1>=l; // case when "=" is on end
    let c2 = i*3+2>=l; // case when "=" is on end
    let chunk = bin(arr[3*i]) + bin(c1? 0:arr[3*i+1]) + bin(c2? 0:arr[3*i+2]);
    let r = chunk.match(/.{1,6}/g).map((x,j)=> j==3&&c2 ? '=' :(j==2&&c1 ? '=':abc[+('0b'+x)]));  
    result += r.join('');
  }

  return result;
}



// TEST

const pic = [ // PNG binary data
    0x89, 0x50, 0x4e, 0x47, 0x0d, 0x0a, 0x1a, 0x0a, 0x00, 0x00, 0x00, 0x0d,
    0x49, 0x48, 0x44, 0x52, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x10,
    0x08, 0x06, 0x00, 0x00, 0x00, 0x1f, 0xf3, 0xff, 0x61, 0x00, 0x00, 0x00,
    0x01, 0x73, 0x52, 0x47, 0x42, 0x00, 0xae, 0xce, 0x1c, 0xe9, 0x00, 0x00,
    0x01, 0x59, 0x69, 0x54, 0x58, 0x74, 0x58, 0x4d, 0x4c, 0x3a, 0x63, 0x6f,
    0x6d, 0x2e, 0x61, 0x64, 0x6f, 0x62, 0x65, 0x2e, 0x78, 0x6d, 0x70, 0x00,
    0x00, 0x00, 0x00, 0x00, 0x3c, 0x78, 0x3a, 0x78, 0x6d, 0x70, 0x6d, 0x65,
    0x74, 0x61, 0x20, 0x78, 0x6d, 0x6c, 0x6e, 0x73, 0x3a, 0x78, 0x3d, 0x22,
    0x61, 0x64, 0x6f, 0x62, 0x65, 0x3a, 0x6e, 0x73, 0x3a, 0x6d, 0x65, 0x74,
    0x61, 0x2f, 0x22, 0x20, 0x78, 0x3a, 0x78, 0x6d, 0x70, 0x74, 0x6b, 0x3d,
    0x22, 0x58, 0x4d, 0x50, 0x20, 0x43, 0x6f, 0x72, 0x65, 0x20, 0x35, 0x2e,
    0x34, 0x2e, 0x30, 0x22, 0x3e, 0x0a, 0x20, 0x20, 0x20, 0x3c, 0x72, 0x64,
    0x66, 0x3a, 0x52, 0x44, 0x46, 0x20, 0x78, 0x6d, 0x6c, 0x6e, 0x73, 0x3a,
    0x72, 0x64, 0x66, 0x3d, 0x22, 0x68, 0x74, 0x74, 0x70, 0x3a, 0x2f, 0x2f,
    0x77, 0x77, 0x77, 0x2e, 0x77, 0x33, 0x2e, 0x6f, 0x72, 0x67, 0x2f, 0x31,
    0x39, 0x39, 0x39, 0x2f, 0x30, 0x32, 0x2f, 0x32, 0x32, 0x2d, 0x72, 0x64,
    0x66, 0x2d, 0x73, 0x79, 0x6e, 0x74, 0x61, 0x78, 0x2d, 0x6e, 0x73, 0x23,
    0x22, 0x3e, 0x0a, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x3c, 0x72, 0x64,
    0x66, 0x3a, 0x44, 0x65, 0x73, 0x63, 0x72, 0x69, 0x70, 0x74, 0x69, 0x6f,
    0x6e, 0x20, 0x72, 0x64, 0x66, 0x3a, 0x61, 0x62, 0x6f, 0x75, 0x74, 0x3d,
    0x22, 0x22, 0x0a, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20,
    0x20, 0x20, 0x20, 0x78, 0x6d, 0x6c, 0x6e, 0x73, 0x3a, 0x74, 0x69, 0x66,
    0x66, 0x3d, 0x22, 0x68, 0x74, 0x74, 0x70, 0x3a, 0x2f, 0x2f, 0x6e, 0x73,
    0x2e, 0x61, 0x64, 0x6f, 0x62, 0x65, 0x2e, 0x63, 0x6f, 0x6d, 0x2f, 0x74,
    0x69, 0x66, 0x66, 0x2f, 0x31, 0x2e, 0x30, 0x2f, 0x22, 0x3e, 0x0a, 0x20,
    0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x3c, 0x74, 0x69, 0x66,
    0x66, 0x3a, 0x4f, 0x72, 0x69, 0x65, 0x6e, 0x74, 0x61, 0x74, 0x69, 0x6f,
    0x6e, 0x3e, 0x31, 0x3c, 0x2f, 0x74, 0x69, 0x66, 0x66, 0x3a, 0x4f, 0x72,
    0x69, 0x65, 0x6e, 0x74, 0x61, 0x74, 0x69, 0x6f, 0x6e, 0x3e, 0x0a, 0x20,
    0x20, 0x20, 0x20, 0x20, 0x20, 0x3c, 0x2f, 0x72, 0x64, 0x66, 0x3a, 0x44,
    0x65, 0x73, 0x63, 0x72, 0x69, 0x70, 0x74, 0x69, 0x6f, 0x6e, 0x3e, 0x0a,
    0x20, 0x20, 0x20, 0x3c, 0x2f, 0x72, 0x64, 0x66, 0x3a, 0x52, 0x44, 0x46,
    0x3e, 0x0a, 0x3c, 0x2f, 0x78, 0x3a, 0x78, 0x6d, 0x70, 0x6d, 0x65, 0x74,
    0x61, 0x3e, 0x0a, 0x4c, 0xc2, 0x27, 0x59, 0x00, 0x00, 0x00, 0xf9, 0x49,
    0x44, 0x41, 0x54, 0x38, 0x11, 0x95, 0x93, 0x3d, 0x0a, 0x02, 0x41, 0x0c,
    0x85, 0xb3, 0xb2, 0x85, 0xb7, 0x10, 0x6c, 0x04, 0x1b, 0x0b, 0x4b, 0x6f,
    0xe2, 0x76, 0x1e, 0xc1, 0xc2, 0x56, 0x6c, 0x2d, 0xbc, 0x85, 0xde, 0xc4,
    0xd2, 0x56, 0xb0, 0x11, 0xbc, 0x85, 0x85, 0xa0, 0xfb, 0x46, 0xbf, 0xd9,
    0x30, 0x33, 0x88, 0x06, 0x76, 0x93, 0x79, 0x93, 0xf7, 0x92, 0xf9, 0xab,
    0xcc, 0xec, 0xd9, 0x7e, 0x7f, 0xd9, 0x63, 0x33, 0x8e, 0xf9, 0x75, 0x8c,
    0x92, 0xe0, 0x34, 0xe8, 0x27, 0x88, 0xd9, 0xf4, 0x76, 0xcf, 0xb0, 0xaa,
    0x45, 0xb2, 0x0e, 0x4a, 0xe4, 0x94, 0x39, 0x59, 0x0c, 0x03, 0x54, 0x14,
    0x58, 0xce, 0xbb, 0xea, 0xdb, 0xd1, 0x3b, 0x71, 0x75, 0xb9, 0x9a, 0xe2,
    0x7a, 0x7d, 0x36, 0x3f, 0xdf, 0x4b, 0x95, 0x35, 0x09, 0x09, 0xef, 0x73,
    0xfc, 0xfa, 0x85, 0x67, 0x02, 0x3e, 0x59, 0x55, 0x31, 0x89, 0x31, 0x56,
    0x8c, 0x78, 0xb6, 0x04, 0xda, 0x23, 0x01, 0x01, 0xc8, 0x8c, 0xe5, 0x77,
    0x87, 0xbb, 0x65, 0x02, 0x24, 0xa4, 0xad, 0x82, 0xcb, 0x4b, 0x4c, 0x64,
    0x59, 0x14, 0xa0, 0x72, 0x40, 0x3f, 0xbf, 0xe6, 0x68, 0xb6, 0x9f, 0x75,
    0x08, 0x63, 0xc8, 0x9a, 0x09, 0x02, 0x25, 0x32, 0x34, 0x48, 0x7e, 0xcc,
    0x7d, 0x10, 0xaf, 0xa6, 0xd5, 0xd2, 0x1a, 0x3d, 0x89, 0x38, 0xf5, 0xf1,
    0x14, 0xb4, 0x69, 0x6a, 0x4d, 0x15, 0xf5, 0xc9, 0xf0, 0x5c, 0x1a, 0x61,
    0x8a, 0x75, 0xd1, 0xe8, 0x3a, 0x2c, 0x41, 0x5d, 0x70, 0x41, 0x20, 0x29,
    0xf9, 0x9b, 0xb1, 0x37, 0xc5, 0x4d, 0xfc, 0x45, 0x84, 0x7d, 0x08, 0x8f,
    0x89, 0x76, 0x54, 0xf1, 0x1b, 0x19, 0x92, 0xef, 0x2c, 0xbe, 0x46, 0x8e,
    0xa6, 0x49, 0x5e, 0x61, 0x89, 0xe4, 0x05, 0x5e, 0x4e, 0xa4, 0x5c, 0x10,
    0x6e, 0x9f, 0xfc, 0x5b, 0x00, 0x00, 0x00, 0x00, 0x49, 0x45, 0x4e, 0x44,
    0xae, 0x42, 0x60, 0x82
];

let b64pic = bytesArrToBase64(pic);
myPic.src = "data:image/png;base64,"+b64pic;
msg.innerHTML = "Base64 encoded pic data:<br>" + b64pic;
img { zoom: 10; image-rendering: pixelated; }
#msg { word-break: break-all; }
<img id="myPic">
<code id="msg"></code>
Kamil Kiełczewski
  • 53,729
  • 20
  • 259
  • 241
1

When I use

btoa("☸☹☺☻☼☾☿"))

I get Error InvalidCharacterError: The string to be encoded contains characters outside of the Latin1 range.

I found documentation

https://developer.mozilla.org/en-US/docs/Web/API/WindowOrWorkerGlobalScope/btoa#unicode_strings

was provided solution as below

function toBinary(string) {
  const codeUnits = new Uint16Array(string.length);
  for (let i = 0; i < codeUnits.length; i++) {
    codeUnits[i] = string.charCodeAt(i);
  }
  return String.fromCharCode(...new Uint8Array(codeUnits.buffer));
}

function fromBinary(binary) {
  const bytes = new Uint8Array(binary.length);
  for (let i = 0; i < bytes.length; i++) {
    bytes[i] = binary.charCodeAt(i);
  }
  return String.fromCharCode(...new Uint16Array(bytes.buffer));
}

const myString = "☸☹☺☻☼☾☿"
// console.log(btoa(myString)) // Error InvalidCharacterError: The string to be encoded contains characters outside of the Latin1 range.
const converted = toBinary(myString)
const encoded = btoa(converted)
console.log(encoded)

const decoded = atob(encoded)
const original = fromBinary(decoded)
console.log(original);
Carson
  • 1,762
  • 1
  • 10
  • 23